Alzheimer's disease (AD), the most common form of dementia among elderly population (prevalence: 1000/100,000; >65 years), represents the fourth leading cause of death in the developed world. Cortical atrophy, neuronal loss, region-specific amyloid deposition, neuritic plaques, and neurofibrillary tangles are key neuropathological features in the AD brain. These alterations are thought to be linked to cognitive decline which clinically defines AD. Within these markers, neuritic plaques are amyloid immunoreactive, thioflavin positive, and accompanied by astrogliosis, microgliosis, cytoskeletal changes, and synaptic loss. The degree of neuritic degeneration within plaques correlates with clinical parameters of dementia. Neuritic plaques are spherical, multicellular lesions that are usually found in moderate to large numbers in limbic structures and associated neocortices of the AD brain. These plaques are comprised of extracellular deposits of amyloid-β peptide(s) (Aβ) that include abundant amyloid fibrils intermixed with non-fibrillary forms of the peptide. Such plaques also contain variable numbers of activated microglia that are often situated very near the fibrillar amyloid core, as well as reactive astrocytes surrounding the core.
The major constituent of the neuritic plaque, β-amyloid polypeptide (Aβ), arises from a larger precursor protein, the amyloid precursor protein (APP) (Kang, et al., 1987; Tanzi, et al., 1987). Aβ is produced by normal cells and can be detected as a circulating peptide in the plasma and cerebrospinal fluid (CSF) of healthy humans. Although the physiological role of the amyloid precursor protein (APP) in the brain is not well understood, missense mutations in APP confer autosomal dominant inheritance of AD (FAD), and shed light on potentially important pathogenic mechanism(s). The accumulation of Aβ, a 39-42 amino acid proteolytic product of APP, in neuritic plaques, structures which at autopsy fulfill the neuropathological criteria for a definitive diagnosis of AD, is thought to be causative for disease progression. A major Aβ cleavage product of APP is the Aβ(1-42) polypeptide, but Aβ peptides shorter at the C-terminus (39 to 41) are also produced by the proteolytic (γ-secretase) cleavage in the membrane. The N-terminal part of Aβ(1-42) is localized in the extracellular region of APP, and the major C-terminal part of the Aβ peptide is contained within the transmembrane domain.
Missense mutations, in APP associated with FAD, occur in proximity to the Aβ domain and result in an increase in the production of the 4 kDa Aβ peptide. In AD, it has been postulated that increased synthesis and/or a decreased clearance of Aβ may lead to amyloid plaque deposition and subsequently to the neuropathological changes associated with the disease. In vitro studies, using synthetic Aβ peptide(s), have shown that neurotoxicity is dependent on Aβ being fibrillar and predominantly in a β-pleated sheet conformation.
The accumulation of extracellular plaques containing the neurotoxic amyloid peptide fragment (Aβ) of β-amyloid precursor protein (APP), as the major product, is one of the characteristics of Alzheimer's disease (AD). Although APP has been recognized as a key molecule for AD, the molecular (patho-) physiological degradation and proteolytic pathways of APP, and cellular interactions and biochemical fate of Aβ peptide(s) are still unclear. Despite the lack of details on degradation pathways and cellular transport for the formation and deposition of Aβ-derived plaques, recent studies towards the development of immunisation methods of AD based on therapeutically active antibodies produced from Aβ(1-42) have yielded initial success in transgenic mouse models of Alzheimer's disease. Several reports have demonstrated that antibodies generated by immunization with Aβ(1-42) are capable of inhibiting the formation of Aβ-plaques by disaggregating Aβ-fibrils, and improve the impairments in the spatial memory of mice. The transgenic APPV717F mouse (TG mouse) is a well characterized model of AD-like plaque pathology with age- and region-dependent deposits of Aβ(1-40) and Aβ(1-42) (Games, et al., 1995). Recently, Schenk et al. and others investigated alterations in the deposition of Aβ in APPV717F TG mouse following immunization with pre-aggregated Aβ(1-42) or administration of antibodies against Aβ (Bard, et al., 2000; Schenk, et al., 1999). Both immunization and administration of Aβ antibodies significantly attenuated amyloid plaque deposition, neuritic dystrophy, and astrogliosis. In these studies, increased titers of mouse anti-human Aβ-antibodies were necessary for the observed reduction in plaque burden. These findings raise the possibility that formation and clearance of an Aβ-antigen: antibody complex may decrease brain Aβ deposition either following antibody generation within the central nervous system or by peripheral antibody transport across the blood-brain-barrier (BBB). Furthermore, passive immunization appears to reduce brain Aβ burden by altering Aβ equilibrium between the CNS and plasma (DeMattos, et al., 2001). Remarkably, active or passive immunization significantly reverses behavioral and memory impairment in APPV717F mouse or other APP transgenic mice (Dodart, et al., 2002; Janus, et al., 2000; Morgan, et al., 2000). These results suggest that immunization may prevent memory deficits possibly by altering a soluble pool of Aβ. Thus, treatment of AD patients with active or passive immunization is one of several emerging therapeutic approaches targeting the production, clearance, and aggregation of the Aβ peptide.
Based on these results, a clinical trial using an active immunization procedure [Aβ(1-42) peptide and/or preaggregates thereof, adjuvant: QS21] was initiated for treatment of patients with established AD. Unfortunately, severe side-effects developed (“meningoencephalitis”) and the clinical trial was stopped. A subgroup of AD patients (n=30) treated with active immunization in this clinical trial, was analyzed (Hock, et al., 2002; Hock, et al., 2003). The authors demonstrated that (i) immunization induces the production of antibodies against Aβ(1-42) and (ii) in patients where a production of antibodies was observable, the cognitive decline was significantly reduced in comparison to the untreated control group. The authors concluded that immunization may be a therapeutic option for AD.
Recent studies elucidated in more detail the recognition properties of antibodies produced upon immunization with Aβ(1-42). This work resulted in the identification of a specific Aβ-epitope recognized by the antibodies generated in transgenic AD mice (McLaurin et al., 2002; Przybylski et al., 2003). These results have been obtained by using selective proteolytic excision technologies (Epitope-Excision) in combination with high resolution mass spectrometry (FTICR-MS) as bioanalytical tools of high sensitivity and specificity for the identification of antigen epitopes (Macht et al 1996; Suckau et al 1992; Macht et al. 2004; see FIGS. 1, 2)). Using mass spectrometric epitope excision of the immobilized Aβ-antigen-immune complex, the epitope was identified to consist of the residues (4-10) (FRHDSGY) of Aβ(1-42). The selectivity of this recognition structure was ascertained by elucidation of the identical epitope from AD plaques, Aβ(1-42) extracts from Aβ-protofibrils, chemically synthesised Aβ(1-42), and other (Aβ-independent) polypeptides comprising the N-terminal Aβ sequence (Przybylski et al. 2003).
Naturally occurring anti-Aβ autoantibodies (Aβ-autoantibodies) were identified by Du et al. in both the blood and the CSF from non-immunized humans (Du, et al., 2001). These antibodies specifically recognize human Aβ as has been shown by immunoprecipitation (Du, et al., 2001) and ELISA. Furthermore, the antibodies readily recognize synthetic Aβ(1-40) as well as human Aβ deposited in the brain of PDAPP transgenic mice. In addition, fibrillation/oligomerization and neurotoxicity of Aβ-peptides were reduced in the presence of Aβ-autoantibodies (Du, et al., 2003).
Furthermore, it has been investigated whether there is a difference of the Aβ-autoantibody concentration in patients with Alzheimer's disease compared to controls. Interestingly, a significant difference among the two groups was found, resulting in a substantially decreased titer (approximately 15-20-fold) of antibodies against Aβ in patients with Alzheimer's disease. These results have been confirmed recently by other groups (Weksler et al., 2004). Antibodies against Aβ can also be detected in commercially available intravenous IgG preparations (IVIgG). The treatment of patients with different neurological diseases with these intravenous immunoglobulin preparations led to the reduction of Aβ concentration in the CSF (Dodel, et al., 2002). The substantial effect of the Aβ-autoantibodies in preventing, and protecting against Aβ-plaque deposition was also established in young (4 months) APP-transgenic (TgCRND8) mice. Additionally, in a pilot trial with 5 patients with AD, utilizing IVIgG, total Aβ was reduced significantly in the CSF and increased in the serum upon delivery of IVIgG (Dodel, et al., 2004). In the five investigated patients no cognitive deterioration was observed during the six months observation period. These results have been confirmed by a recent pilot study involving 8 AD patients, who were treated with IVIgG (Relkin et al., 2006)
However, administering IVIgG to a patient with AD is not convenient and associated with high costs, as the fraction of therapeutic Aβ autoantibodies is low. The vast majority of IgG in this preparation is not Aβ specific and may result in undesirable effects. Furthermore, the sources for IVIgG are limited, which is an unacceptable disadvantage in view of the prevalence of patients with Alzheimer's disease.
Methods of detecting and monitoring the progression AD and other neurodementing diseases similarly are inadequate. Current AD diagnostics fall into three groups: (i) determinations for genetic risk factors or mutations (mainly for FAD cases, but not for sporadic AD diagnostics); (ii) neuroimaging methods; and (iii) diagnostics based on biochemical/biological markers. Present work on the development of diagnostic procedures based on biomarkers have been mainly focused on CSF, which has the principal disadvantage that such methods require elaborate, invasive material. A major problem associated with brain-derived biomarkers is that clinically examined controls often also include subjects with preclinical AD pathology. Further, current available biomarkers have the major disadvantage of low specificity. Similar disadvantages have been noted for a series of proteins expressed in the frontal cortex, identified by brain proteomics approaches, as potential brain biomarkers arising from presumed alterations of blood brain barrier in AD.
Studies on biomarkers in plasma and serum have been performed mainly with determinations of SP (senile plaques) and NFT (neurofibrillary tangles) components, e.g. the Aβ peptides Aβ(1-40) (SEQ ID NO: 1) and Aβ(1-42) (found with elevated levels) and hyperphosphorylated Tau-protein. However the specificity of Aβ determinations, and application for early and differential diagnostics has been considered uncertain, the same is the case for protein Tau determinations which has been described as a marker of already progressing neurodegeneration.
There exists, therefore, a need for improved methods of treating and detecting neurodementing diseases such as AD.